Numerous studies have been published describing the effect of various probiotics (PRO) on behaviours related to psychiatric disease. We have previously shown a robust antidepressant-like effect of PRO in rats, but over time, the treatment effect seems to vary significantly between different sets of rats from the same commercial vendor. Therefore, we hypothesised that the antidepressant-like response may be modulated by the cohabiting gut microbiota. The aims of the present study were (1) to investigate any differences in the gut microbiota composition between responders (Resp) and non-responders (Non-resp) to PRO with regards to depressive-like behaviour, and (2) to evaluate the effects of PRO on the microbiota composition. Two sets of 20 male Sprague-Dawley rats each were treated with multi-species PRO (nine Bifidobacterium, Lactococcus and Lactobacillus species) for eight weeks and subjected to a behavioural assessment. Faecal samples were collected for 16 s rRNA (VR4) gene amplicon sequencing (Illumina MiSeq). As previously reported, PRO-treated Resp animals showed a marked decrease in depressive-like behaviour, whereas no such response was seen in Non-resp. We observed profound differences in the gut microbiota composition between the two sets of rats, and the relative faecal abundance of the genera that comprised PRO was higher in Resp than in Non-resp although treated with the same dose of PRO. Particularly, the relative abundance of the Lactobacillus genus was not increased in PRO-treated Non-resp animals. In conclusion, the cohabiting microbiota and the faecal abundance of PRO may modulate the antidepressant-like effect of PRO in rats.